CN107197252B - Decoding device for performing intra prediction - Google Patents

Abstract

The present invention relates to an intra prediction method and apparatus thereof. The method for intra prediction according to the present invention comprises the steps of: determining a first MPM candidate corresponding to a left candidate block on the basis of a position of the left candidate block adjacent to the left of the current block; determining a second MPM candidate corresponding to an upper candidate block on the basis of a position of the upper candidate block adjacent to an upper end of the current block; generating an MPM candidate list including a plurality of MPM candidates on the basis of the first MPM candidate and the second MPM candidate; and determining one MPM candidate of a plurality of MPM candidates constituting the MPM candidate list as an intra prediction mode of the current block, and performing intra prediction on the current block on the basis of the determined intra prediction mode to generate a prediction block corresponding to the current block.

Description

Decoding apparatus performing intra prediction

This application was submitted on June 11, 2014, the international application date is October 16, 2012, the application number is 201280061064.1 (PCT/KR2012/008442), and the name of the invention is "Intra-frame prediction method and device" patent Divisional application of the application.

technical field

The present invention relates to image processing, and more particularly, to intra prediction methods and apparatuses.

Background technique

Recently, in various fields, demands for high-resolution and high-quality images such as high-definition (HD) images and ultra-high-definition (UHD) images are increasing. Compared with existing image data, high-resolution and high-quality image data requires a relatively large amount of information or bits. Therefore, when image data is transmitted using a medium such as an existing wired/wireless broadband line, or when image data is stored by using an existing storage medium, the transmission cost and storage cost increase. To solve these problems, efficient image compression techniques can be used.

Regarding image compression techniques, there are various techniques, such as inter prediction techniques, which predict pixel values included in the current picture from previous and/or next pictures of the current picture; intra prediction techniques, which use Pixel information predictions include pixel values in the current picture; entropy coding techniques, which assign short codewords to values with high frequency of occurrence and long codewords to values with low frequency of occurrence, etc. This image compression technique can be used to transmit or store image data by efficiently compressing the image data.

SUMMARY OF THE INVENTION

technical problem

The present invention provides an image encoding method and an apparatus capable of improving encoding/decoding efficiency.

The present invention also provides an image decoding method and an apparatus capable of improving encoding/decoding efficiency.

The present invention also provides an intra prediction method and an apparatus capable of improving encoding/decoding efficiency.

The present invention also provides an intra prediction mode derivation method and a device capable of improving encoding/decoding efficiency.

Technical solutions

According to an aspect of the present invention, an intra prediction method is provided. The method includes: determining a first most probable mode (MPM) candidate corresponding to the left candidate block on the basis of the position of the left candidate block adjacent to the left side of the current block; On the basis of the position of the upper candidate block, determine the second MPM candidate corresponding to the upper candidate block; On the basis of the first MPM candidate and the second MPM candidate, generate an MPM candidate list including a plurality of MPM candidates; And the One MPM candidate is determined as the intra prediction mode of the current block among the plurality of MPM candidates constituting the MPM candidate list, and intra prediction is performed on the current block on the basis of the determined intra prediction mode, so as to generate an intra prediction mode corresponding to the current block. prediction block.

In the above aspect of the present invention, the first MPM candidate and At least one MPM candidate between the second MPM candidates.

Also, if the upper candidate is located outside the CTB to which the current block belongs, the determination of the first MPM candidate may further include assigning a specific intra prediction mode to the upper candidate block, and determining the intra prediction mode assigned to the upper candidate block is the first MPM candidate.

Also, the specific intra prediction mode may be the DC mode.

In addition, in the determination of the first MPM candidate, if the left candidate block is located outside the current picture to which the current block belongs, the first MPM candidate may be determined as the DC mode, and in the determination of the second MPM candidate, if the above If the candidate block is located outside the current picture, the second MPM candidate can be determined to be the DC mode.

In addition, in the determination of the first MPM candidate, if the left candidate block is located outside the current segment to which the current block belongs, the first MPM candidate may be determined as the DC mode, and in the determination of the second MPM candidate, if the above The candidate block is located outside the current segment, then the second MPM candidate can be determined as DC mode.

In addition, in the determination of the first MPM candidate, if the prediction mode of the left candidate block is not the intra mode, the first MPM candidate may be determined as the DC mode, and in the determination of the second MPM candidate, if the prediction mode of the upper candidate block is If the prediction mode is not the intra mode, the second MPM candidate may be determined as the DC mode.

According to another aspect of the present invention, an image decoding method is provided. The method includes: on the basis of the position of the left candidate block adjacent to the left side of the current block, determining a first MPM candidate corresponding to the left candidate block; On the basis of the position, determine the second MPM candidate corresponding to the upper candidate block; generate an MPM candidate list including a plurality of MPM candidates on the basis of the first MPM candidate and the second MPM candidate; determine one MPM candidate as constituting the MPM an intra prediction mode of the current block in the plurality of MPM candidates of the candidate list, and performing intra prediction on the current block on the basis of the determined intra prediction mode to generate a prediction block corresponding to the current block; and Reconstruction blocks are generated on the basis of blocks.

In the above aspect of the present invention, the first MPM candidate and the second MPM candidate may be determined according to whether the candidate block corresponding to at least one MPM candidate between the left candidate block and the upper candidate block is located outside the CTB to which the current block belongs. at least one MPM candidate in between.

Also, if the upper candidate is located outside the CTB to which the current block belongs, the determination of the first MPM candidate may further include assigning a specific intra prediction mode to the upper candidate block, and assigning an intra prediction mode assignable to the upper candidate block Determined to be the first MPM candidate.

Also, the specific intra prediction mode may be the DC mode.

In addition, in the determination of the first MPM candidate, if the left candidate block is located outside the current picture to which the current block belongs, the first MPM candidate may be determined as the DC mode, and in the determination of the second MPM candidate, if the above If the candidate block is located outside the current picture, the second MPM candidate can be determined to be the DC mode.

In addition, in the determination of the first MPM candidate, if the left candidate block is located outside the current segment to which the current block belongs, the first MPM candidate may be determined as the DC mode, and in the determination of the second MPM candidate, if the above If the candidate block is located outside the current segment, the second MPM candidate may be determined as the DC mode.

Also, in the determination of the first MPM candidate, if the prediction mode of the left candidate block is not the intra mode, the first MPM candidate may be determined as the DC mode, and in the determination of the second MPM candidate, if the prediction mode of the upper candidate block is If the prediction mode is not the intra mode, the second MPM candidate may be determined as the DC mode.

According to another aspect of the present invention, an image decoding apparatus is provided. The apparatus includes: an intra-frame predictor for determining an intra-frame prediction mode of a current block and performing intra-frame prediction on the current block on the basis of the determined intra-frame prediction mode, so as to generate a prediction block corresponding to the current block ; and a reconstructed block generator for generating reconstructed blocks based on the predicted blocks. Here, the intra predictor may determine the first MPM candidate corresponding to the left candidate block on the basis of the position of the left candidate block adjacent to the left side of the current block, and the first MPM candidate corresponding to the upper part of the current block is adjacent to the current block. On the basis of the position of the upper candidate block, the second MPM candidate corresponding to the upper candidate block is determined, and an MPM candidate list including multiple MPM candidates can be generated on the basis of the first MPM candidate and the second MPM candidate, and the One MPM candidate is determined as the intra prediction mode of the current block among a plurality of MPM candidates constituting the MPM candidate list.

[Beneficial effect]

The image encoding method according to the present invention can improve image encoding/decoding efficiency.

The image decoding method according to the present invention can improve image encoding/decoding efficiency.

The intra prediction method according to the present invention can improve image encoding/decoding efficiency.

The intra prediction mode derivation method of the present invention can improve image encoding/decoding efficiency.

Description of drawings

FIG. 1 is a block diagram of an image encoder according to an embodiment of the present invention.

Figure 2 is a schematic diagram of the concept of a predictor according to an embodiment of the present invention.

FIG. 3 is a block diagram of an image decoder according to an embodiment of the present invention.

FIG. 4 is a schematic diagram illustrating a concept of a predictor of an image decoder according to an embodiment of the present invention.

5 is a schematic diagram illustrating a concept of an example of a quad-tree structure of processing units in a system according to the present invention.

FIG. 6 is a flowchart illustrating a method of transmitting intra prediction mode information according to an embodiment of the present invention.

FIG. 7 is a flowchart illustrating a method of deriving an intra prediction mode according to an embodiment of the present invention.

FIG. 8 is a schematic diagram illustrating an embodiment of neighboring blocks for deriving the most probable mode (MPM) candidates.

FIG. 9 is a diagram explaining a line buffer in which an intra prediction mode is stored.

FIG. 10 is a schematic diagram illustrating an embodiment of a method of deriving MPM candidates for blocks adjacent to a boundary of a largest coding unit (LCD).

FIG. 11 is a schematic diagram illustrating an embodiment of a method of deriving MPM candidates according to the present invention.

FIG. 12 is a schematic diagram illustrating an embodiment of a method of deriving MPM candidates on the basis of intra mode storage units.

13 is a schematic diagram illustrating an embodiment of a 2:1 line buffer compression scheme.

14 is a schematic diagram illustrating an embodiment of a 4:1 line buffer compression scheme.

Detailed ways

As the invention is possible with various modifications and various embodiments, only specific embodiments are shown by way of example in the drawings and are described in detail below. It should be understood, however, that the invention is not limited to the specific embodiments described herein. The terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the invention. As used herein, the singular form is also intended to include the plural form unless the context clearly dictates otherwise. In this application, it should be understood that terms such as "including" or "having" are intended to refer to features, numbers, operations, acts, components, parts or combinations thereof disclosed in this specification The presence of, and is not intended to preclude the presence or addition of, one or more other features, numbers, operations, acts, components, parts, or combinations thereof.

Meanwhile, the corresponding configurations in the drawings described in the present invention are independently shown to facilitate explanation about different specific functions in the image encoder/decoder, but do not mean to utilize a separate hardware entity or a separate software entity Perform the corresponding construction. For example, in respective configurations, two or more configurations may be combined into one configuration, and one configuration may be divided into multiple configurations. Embodiments in which corresponding configurations are integrated and/or separate are also included within the scope of the present invention, as long as they do not depart from the spirit of the present invention.

Furthermore, some components may not be essential components for performing inherent functions, but optional components only for improving performance. The present invention can be implemented by including only the basic constituent elements for carrying out the spirit of the present invention, except for the constituent elements only for improving the formation. Structures in which only essential constituents are included in addition to optional constituents only for improving performance are also included in the scope of the present invention.

Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings. Also, throughout the drawings, the same reference numerals are used to designate the same elements, and the same descriptions about the same elements will be omitted.

FIG. 1 is a block diagram of an image encoder according to an embodiment of the present invention. 1, the image encoder 100 includes a picture divider 105, a predictor 110, a transformer 115, a quantizer 120, a rearranger 125, an entropy encoder 130, a dequantizer 135, an inverse transformer 140, a filter 145, and memory 150.

The picture divider 105 divides the input picture on the basis of at least one processing unit. In this case, the processing unit may be a prediction unit (PU), a transform unit (TU), or a coding unit (CU).

The predictor 110 as described below may include an inter-frame predictor that performs inter-frame prediction and an intra-frame predictor that performs intra-frame prediction. The predictor 110 may generate prediction blocks by performing prediction on the processing units of the pictures in the picture divider 105. The processing unit of a picture in predictor 100 may be a CU, TU or PU. In addition, it is determined whether the prediction performed for the corresponding processing unit is inter prediction or intra prediction, and the specific content (eg, prediction mode, etc.) of each prediction method can be determined. In this case, the processing unit for performing prediction and the processing unit for determining specific content may be different. For example, a prediction method, prediction mode, etc. may be determined in a PU unit, and prediction may be performed in a TU unit. Residual values (ie, residual blocks) between the generated prediction block and the original block may be input to the transformer 115 . Furthermore, prediction mode information, motion vector information, etc. for prediction may be encoded in the entropy encoder 130 together with the residual values, and may be passed to the decoder.

The transformer 115 transforms the residual block on a transform unit basis, and generates transform coefficients. The transform unit of the transformer 115 may be a TU, and may have a quad-tree structure. In this case, the size of the transform unit may be determined within a certain maximum or minimum size range. Transformer 115 may transform the residual block by using discrete cosine transform (DCT) and/or discrete sine transform (DST).

The quantizer 120 may generate quantized coefficients by quantizing the residual values transformed in the transformer 115 . The value calculated by the quantizer 120 may be provided to the dequantizer 135 and the rearranger 125.

The rearranger 125 may rearrange the quantized coefficients provided from the quantizer 120 . The reordering of the quantized coefficients may increase the encoding efficiency of the entropy encoder 130 . The rearranger 125 can rearrange the quantized coefficients in the form of 2-dimensional blocks into the form of 1-dimensional vectors by using the coefficient scanning method. Based on the probability statistics of the coefficients transmitted from the quantizer, the rearranger 125 may change the order in which the coefficients are scanned, thereby increasing the entropy encoding efficiency of the entropy encoder 130.

The entropy encoder 130 may perform entropy encoding on the quantized coefficients rearranged by the rearranger 125 . The entropy encoder 130 may encode various information passed from the rearranger 125 and the predictor 110 . The information may include quantization coefficient information and block type information of coding units, prediction mode information, division unit information, prediction unit information and transmission unit information, motion vector information, reference picture information, interpolation information of blocks, filtering information, and the like.

Entropy coding may use Exponential Golomb, CAVLC (Context Adaptive Variable Length Coding) and/or CABAC (Context Adaptive Binary Algorithm Coding). For example, a table for performing entropy encoding, such as a variable length coding (VCL) table, may be stored in the entropy encoder 130. The entropy encoder 130 may perform entropy encoding by using the stored VLC table. For another example, in the CABAC entropy encoding method, the entropy encoder 130 may convert symbols into binary form (ie, bins), and thus a bitstream may be generated by performing arithmetic encoding on the bins according to the bit generation probability.

When entropy coding is applied, an index with a high value and its corresponding short codeword can be assigned to a symbol with a high probability of generation, and an index with a high value and its corresponding long codeword can be assigned to a symbol with a high value and its corresponding long codeword Symbols with low probability of generation. Therefore, the amount of bits used for symbols to be encoded can be reduced, and image compression performance can be improved by entropy encoding.

The dequantizer 135 may dequantize the values quantized by the quantizer 120 . The inverse transformer 140 may inverse transform the value dequantized by the dequantizer 135 . The residual values generated in dequantizer 135 and inverse transformer 140 may be combined with the prediction block predicted by predictor 110, and thus a reconstructed block may be generated.

Filter 145 may apply loop filters to the reconstructed blocks and/or pictures. Loop filters may include deblocking filters, sample adaptive offset (SAO), and/or adaptive loop filters (ALF), among others.

A deblocking filter can remove block distortion that occurs at the boundaries between blocks in the reconstructed picture. SAO can add appropriate offset values to pixel values to compensate for encoding errors. The ALF may perform filtering on the basis of values used to compare the original image with the image reconstructed after filtering the block by a deblocking filter.

Meanwhile, with respect to reconstructed blocks used in intra prediction, the filter 145 may not apply the filtering.

The memory 150 may store reconstructed blocks or pictures calculated by using the filter 145 . The reconstructed blocks or pictures stored in memory 150 may be provided to predictor 110 for performing inter prediction.

FIG. 2 is a schematic diagram illustrating the concept of a predictor according to an embodiment of the present invention. Referring to FIG. 2 , the predictor 200 may include an inter-frame predictor 210 and an intra-frame predictor 220.

The inter predictor 210 may generate a prediction block by performing prediction on the basis of the information of the previous picture or the next picture of the current picture. Regarding a prediction unit (PU), the inter predictor 210 may select a reference picture, and may select a reference block having the same size as the PU as a pixel sampling unit. Subsequently, the inter predictor 210 may generate prediction blocks that are sampling units smaller than integer units (eg, 1/2 pixel sampling unit and 1/4 pixel sampling unit), and thus most similar to the current PU, and the residual signal therein The motion vector that can be minimized and to be encoded can also be minimized. In this case, the motion vector can be expressed in units smaller than an integer number of pixels.

Also, on the basis of pixel information in the current picture, the intra predictor 220 may generate a prediction block by performing prediction. In this case, regarding the PU, the intra predictor 220 may determine an intra prediction mode, and may perform prediction based on the determination of the intra prediction mode.

The index of the reference picture selected by the inter predictor 210, the information on the motion vector, and the information on the intra prediction mode selected by the intra predictor 220 may be encoded and passed to the decoder.

FIG. 3 is a block diagram of an image decoder according to an embodiment of the present invention. Referring to FIG. 3 , the image decoder 300 may include an entropy decoder 310 , a rearranger 315 , a dequantizer 320 , an inverse transformer 325 , a predictor 330 , a filter 335 and a memory 340 .

When the image bitstream is input to the image decoder, the input bitstream can be decoded according to the operation of processing image information in the image encoder.

The entropy decoder 310 may perform entropy decoding on the input bitstream. The entropy decoding method is similar to the entropy encoding method described above. When entropy decoding is applied, an index with a high value and its corresponding short codeword can be assigned to a symbol with a high probability of generation, and an index with a high value and its corresponding long codeword can be assigned to a symbol with a high value and its corresponding long codeword Symbols with low probability of generation. Therefore, the bit amount of symbols to be encoded is reduced, and image compression performance can be improved by entropy encoding.

Among the pieces of information decoded by the entropy decoder 310 , information for generating a prediction block may be provided to the predictor 330 , and residual values subjected to entropy decoding in the entropy decoder may be input to the rearranger 315 .

The rearranger 315 reorders the bitstream subjected to entropy decoding in the entropy decoder 310 according to the reordering method used in the image encoder. The rearranger 315 can perform reordering by reconstructing coefficients expressed in a 1-dimensional vector format into coefficients in a 2-dimensional block format. The rearranger 315 may receive information on coefficient scans performed in the encoder, and may perform the reordering by using an inverse scan method based on the order of scans performed in the encoder.

The dequantizer 320 may perform dequantization on the basis of the quantization parameters provided from the encoder and the coefficient values of the reordered blocks.

According to the quantization result performed by the image encoder, the inverse transformer 325 may perform inverse DCT and/or inverse DST with respect to DCT and DST performed by the transformer of the encoder. The inverse transformation may be performed based on the determination of the transmission unit or the image division unit. The transformer of the encoder may selectively perform DCT and/or DST according to pieces of information such as prediction method, current block size, and/or prediction direction. The inverse transformer 325 of the decoder may perform the inverse transform on the basis of information about the transform performed in the transformer of the encoder.

The predictor 330 may generate the prediction block on the basis of the prediction block generation related information provided from the entropy decoder 310 and the previously decoded block and/or picture information provided from the memory 340 . The reconstructed block may be generated by using the prediction block generated from the predictor 330 and the residual block provided from the inverse transformer 325 .

The reconstructed blocks and/or pictures may be provided to filter 335 . Filter 335 may apply a loop filter to the reconstructed block and/or picture. Loop filters may include deblocking filters, sample adaptive offset (SAO), and/or adaptive loop filters (ALF), among others.

Memory 340 may store reconstructed pictures or blocks for use as reference pictures or blocks, or to provide reconstructed pictures to output elements.

FIG. 4 is a schematic diagram illustrating a concept of a predictor of an image decoder according to an embodiment of the present invention.

Referring to FIG. 4 , the predictor 400 may include an inter-frame predictor 410 and an intra-frame predictor 420 .

If the prediction mode for the PU is the inter prediction mode, the inter predictor 410 may use information required for inter prediction of the current PU provided from the image encoder, for example, information on motion vectors, reference picture indexes, and the like , to perform inter prediction on the current PU based on information included in at least one of the previous and next pictures of the current picture in which the current PU is included. In this case, if the skip flag, merge flag, etc. of a coding unit (CU) received from the encoder is confirmed, motion information can be derived therefrom.

When the prediction mode for the corresponding PU is the intra prediction mode, the intra predictor 420 may generate a prediction block based on pixel information in the current picture. In this case, regarding the PU, the intra predictor 420 may determine an intra prediction mode, and may perform prediction based on the determined intra prediction mode. Here, if the intra prediction mode related information received from the encoder is confirmed, the intra prediction mode can be derived therefrom.

Here, if 'image' or 'screen' can be used in the same meaning as 'picture' configured or expressed according to the present invention, 'picture' may be referred to as 'image' or 'screen'.

5 is a schematic diagram illustrating a concept of an example of a quad-tree structure of processing units in a system according to the present invention.

A coding unit (CU) may mean a unit that performs encoding/decoding of a picture. One coding block in a picture to be coded based on the quadtree structure may have depth and may be repeatedly divided. In this case, an encoding block that is no longer divided may correspond to a CU, and the encoder may perform encoding processing for the CU. The size of the CU can be different, such as 64x64, 32x32, 16x16, 8x8, etc.

Here, a coding block repeatedly divided based on a quad-tree structure may be referred to as a coding tree block (CTB). One CTB may not be further partitioned, and in this case, the CTB itself may correspond to one CU. Therefore, the CTB may correspond to the largest coding unit (LCU) which is the CU having the largest size. Meanwhile, a CU having the smallest size in the CTB may be referred to as a smallest coding unit (SCU).

Referring to FIG. 5 , with this partitioning, the CTB 500 may have a hierarchical structure composed of smaller CUs 510 . The hierarchical structure of the CTB 500 may be specified based on size information, depth information, segmentation identification information, and the like. Information on the size of the CTB, split depth information, split identification information, etc. may be transmitted from the encoder to the decoder through a sequence parameter set (SPS) included in the bitstream.

Meanwhile, which prediction will be performed between the inter prediction and the intra prediction may be determined in a unit of the CU. If inter prediction is performed, an inter prediction mode, motion information, etc. may be determined in a unit of a PU, and if an intra prediction is performed, an intra prediction mode may be determined in a unit of a PU. In this case, as described above, the processing unit by which the prediction is performed and the processing unit by which the prediction method and its specific content are determined are the same, or the two units may be different. For example, a prediction method, a prediction mode, etc. may be determined in a unit of a PU, and a prediction may be performed in a unit of a transform unit (TU).

Referring to FIG. 5, one CU 510 may function as one PU or may be divided into multiple PUs. In the context of intra prediction 520, the partitioning mode of a CU (and/or PU) may be a 2Nx2N or NxN mode (where N is an integer). Here, a PU in the 2Nx2N mode may have a size of 2Nx2N, and a PU in the NxN mode may have a size of NxN. In the case of inter prediction 530, the partition mode of a CU (and/or PU) may be a 2Nx2N, 2NxN, Nx2N, NxN, 2NxnU, 2NxnD, nLx2N, or nRx2N mode (where N is an integer). Here, a PU in the 2NxN mode may have a size of 2NxN, and a PU in the Nx2N mode may have a size of Nx2N. Also, in the 2NxnU mode, one CU may be partitioned into PUs having a size of 2Nx(1/2)N and PUs having a size of 2Nx(3/2)N. In this case, a PU having a size of 2Nx(1/2)N may be located on top of a PU having a size of 2Nx(3/2)N. In 2NxnD mode, one CU may be partitioned into PUs with a size of 2Nx(3/2)N and PUs with a size of 2Nx(1/2)N. In this case, a PU having a size of 2Nx(1/2)N may be located below a PU having a size of 2Nx(3/2)N. Also, in the nLx2N mode, one CU may be partitioned into a PU having a size of (1/2)Nx2N and a PU having a size of (3/2)Nx2N. In this case, a PU having a size of (1/2)Nx2N may be located to the left of a PU having a size of (3/2)Nx2N. In nRx2N mode, one CU may be partitioned into PUs of size (3/2)Nx2N and PUs of size (1/2)Nx2N. In this case, a PU having a size of (1/2)Nx2N may be located to the right of a PU having a size of (3/2)Nx2N.

The above-described partition mode is only for one embodiment, and thus the method of dividing a CU into PUs is not limited to the above-described embodiment. For example, in the case of inter prediction 530, the partition modes of a CU (and/or PU) may only be used for four types of modes, ie, 2Nx2N, 2NxN, Nx2N, and NxN, and in addition to the above-mentioned eight types of partition modes In addition, another split mode can be used further.

Here, in the present invention, the current block is a block for which encoding, decoding and/or prediction processing is currently performed, and means a block corresponding to a processing unit when encoding, decoding and/or prediction processing is performed. For example, if prediction processing is performed on the current block, the current block may correspond to a block to be predicted corresponding to the current PU. Also, in the present invention, a block generated by prediction is referred to as a prediction block.

A "unit" means a processing unit when encoding, decoding, etc. is performed, and is thus distinguishable from a "block" indicating a group of pixels and/or a group of samples. However, for convenience of explanation, 'unit' may alternatively represent a 'block' corresponding to the 'unit' in the present invention. For example, hereinafter, in the present invention, a block to be predicted corresponding to one PU may be referred to as a PU, and a block to be encoded/decoded and corresponding to one CU may be referred to as a CU. This distinction will be more clearly understood by those skilled in the art.

Meanwhile, as described above in the embodiments of FIGS. 2 and 4 , the intra predictor may perform prediction on the basis of pixel information in the current picture, and thus may generate a prediction block for the current block. For example, the intra predictor may predict pixel values in the current block by using pixels in reconstructed blocks located in upper, left, upper left, and/or upper right portions adjacent to the current block.

The intra prediction mode may be a vertical mode, a horizontal mode, a DC mode, a planar mode, a corner mode according to the position of a reference pixel used to predict the pixel value of the current block and/or a prediction scheme, etc. In the vertical mode, prediction can be performed in the vertical direction by using pixel values of adjacent blocks. In the horizontal mode, prediction can be performed in the horizontal direction by using pixel values of adjacent blocks. Also, in DC mode, the pixel value in the current block can be predicted by using the average pixel value around the current block. In the planar mode, on the basis of pixel values of a plurality of pixels located adjacent to the current block, the predicted value of the pixel located in the current block to be predicted may be derived through specific calculation. In this case, a plurality of pixels for predicting the pixel to be predicted may be differently determined according to the position of the pixel to be predicted. In the angle mode, prediction may be performed according to the predicted angle and/or direction with respect to each mode.

The intra predictor may use predetermined prediction direction and prediction mode values to perform intra prediction. In this case, for example, the number of intra prediction modes assignable to the current block may differ according to the size of the current block. Table 1 below shows an embodiment of the number of intra prediction modes that can be allocated to a current block (and/or PU) according to the size of the current block (and/or PU).

[Table 1]

PU size number of patterns 4x4 18 8x8 35 16x16 35 32x32 35 64x64 4

For another example, the number of intra prediction modes assignable to the current block may be a certain fixed value. For example, the number of intra prediction modes assignable to the current block may be 35. In this case, the 35 intra prediction modes may include the above-described DC, planar, vertical, horizontal, angular modes, and the like.

As described above, after determining the intra prediction mode, the encoder may encode information about the determined intra prediction mode and then transmit it to the decoder. Although the intra prediction mode information may be transmitted as its own value indicating its prediction mode, it is also possible to provide a method of transmitting intra prediction mode information based on a mode value predicted for the intra prediction mode to increase transmission efficiency possible. Hereinafter, the prediction mode used as the prediction value for the intra prediction mode of the current block is referred to as the most probable mode (MPM) in the present invention.

FIG. 6 is a flowchart illustrating a method of transmitting intra prediction mode information according to an embodiment of the present invention.

Referring to FIG. 6, the encoder may derive a plurality of MPM candidates constituting an MPM candidate list on the basis of a plurality of adjacent blocks adjacent to the current block (S610).

The encoder may derive multiple MPM candidates on the basis of multiple adjacent blocks, and the MPM candidate list may be generated by assigning the MPM candidates to the MPM candidate list. In this case, the encoder may directly use the intra prediction mode of the adjacent block as the MPM candidate corresponding to the adjacent block, or may use the specific intra prediction mode determined according to a specific condition as the adjacent block Corresponding MPM candidates.

Meanwhile, the encoder may use a certain fixed number of MPM candidates to encode intra prediction modes. In this case, the number of MPM candidates included in the MPM candidate list may be equal to a certain fixed number. For example, the number of MPM candidates constituting the MPM candidate list may be three. Hereinafter, for convenience of explanation, it is assumed that the number of MPM candidates constituting the MPM candidate list is three in the present invention.

In this case, the number of adjacent blocks for deriving MPM candidates may be smaller than the number of MPM candidates constituting the MPM candidate list. For example, if the number of adjacent blocks used to derive MPM candidates is 2, the number of MPM candidates derived from adjacent blocks may be 2. In this case, since the number of MPM candidates constituting the MPM candidate list is fixed to 3, the encoder can determine additional MPM candidates and can assign them to the MPM candidate list. Here, in addition to the MPM candidates derived from neighboring blocks, during the intra prediction mode, additionally derived MPM candidates may be selected.

A detailed embodiment of a method of deriving MPM candidates based on prediction modes of neighboring blocks will be described below.

Referring back to FIG. 6, the encoder may encode intra prediction mode information based on the MPM candidate list, and may transmit it to the decoder (S620).

The encoder determines whether the same MPM candidate as the intra prediction mode of the current block exists in the multiple MPM candidates constituting the MPM candidate list, that is, whether the predicted value of the intra prediction mode is directly used as the intra prediction mode of the current block. , MPM identification information can be generated. Here, the MPM flag may correspond to a flag indicating whether or not MPM candidates with the same intra prediction mode of the current block exist in a plurality of MPM candidates constituting the MPM candidate list, and may be represented by prev_intra_luma_pred_flag, for example. The generated MPM identification information may be encoded by the encoder's entropy encoder and may then be transmitted to the decoder.

If the same MPM candidate as the intra prediction mode of the current block exists in the current block list, the encoder may generate MPM index information indicating which candidate is the same as the intra prediction mode of the current block among the multiple MPM candidates constituting the MPM candidate list The prediction mode is the same. For example, MPM index information may be represented by mpm_idx. In this case, the generated MPM index information may be encoded by the entropy encoder of the encoder, and may then be transmitted to the decoder.

If the same MPM candidate as the intra prediction mode of the current block does not exist in the MPM candidate list, on the basis of the multiple MPM candidates constituting the MPM candidate list and the intra prediction mode of the current block, the encoder can derive the same The remaining mode corresponding to the intra prediction mode of the current block. In this case, the derived mode values of the remaining modes can be encoded by the entropy encoder of the encoder, and can then be transmitted to the decoder.

FIG. 7 is a flowchart illustrating a method of deriving an intra prediction mode according to an embodiment of the present invention.

Referring to FIG. 7, the decoder may perform decoding by receiving intra prediction mode information from the encoder (S710). The decoding process can be performed by the entropy encoder of the decoder. The intra prediction mode information received from the encoder may be MPM identification information, MPM index information, remaining mode information, and the like.

Referring back to FIG. 7 , the decoder may derive a plurality of MPM candidates constituting the MPM candidate list on the basis of a plurality of adjacent blocks adjacent to the current block ( S720 ). That is, the decoder may derive multiple MPM candidates on the basis of multiple adjacent blocks, and may generate the MPM candidate list by assigning the MPM candidates to the MPM candidate list. A detailed embodiment of a method of deriving MPM candidates based on prediction modes of neighboring blocks will be described below.

Referring back to FIG. 7, the decoder may derive the intra prediction mode of the current block on the basis of the MPM candidate list and the intra prediction mode information (S730).

On the basis of the MPM identification information received from the encoder, the decoder may determine whether the MPM candidate that is the same as the prediction mode of the current block exists in a plurality of MPM candidates constituting the MPM candidate list.

If the same MPM candidate as the prediction mode of the current block exists in the MPM candidate list, the decoder may determine the MPM candidate indicated by the MPM index information as the intra prediction mode of the current block. Since the MPM index information is described above with reference to FIG. 6, a detailed description about the MPM index information will be omitted here. If the same MPM candidate as the prediction mode of the current block does not exist in the MPM candidate list, the decoder may derive the intra prediction mode of the current block based on the MPM list and the remaining modes received from the encoder.

When the intra prediction mode of the current block is derived, the decoder may generate a prediction block corresponding to the current block by performing intra prediction on the current block on the basis of the derived intra prediction mode.

FIG. 8 is a schematic diagram illustrating an embodiment of neighboring blocks for deriving MPM candidates. In the embodiment of FIG. 8, the current block and neighboring blocks adjacent to the current block may be corresponding blocks corresponding to one PU.

Referring to 810 of FIG. 8 , on the basis of the block A 813 located at the uppermost part among the left neighboring blocks located on the left side of the current block and the block B 816 located at the leftmost among the upper neighboring blocks adjacent to the upper side of the current block , the intra predictor can derive MPM candidates corresponding to the current block. In this case, the MPM candidate corresponding to the block A 813 (hereinafter referred to as the MPM candidate A) may be determined as the intra prediction mode of the block A 813 , and the MPM candidate corresponding to the block B 816 (hereinafter referred to as the MPM candidate) B) may be determined as the intra prediction mode of block B 816. However, if neighboring blocks (ie, block A and/or block B) are unavailable or satisfy different specific conditions, MPM candidates (ie, block A and/or block B) corresponding to the neighboring blocks may be determined is a specific intra prediction mode. Detailed examples thereof will be detailed below.

Meanwhile, as described above, the number of intra prediction modes assignable to the current block may be different according to the size of the current block. In this case, the mode value of the intra prediction mode assignable to the neighboring block (ie, block A and/or block B) may be greater than the maximum mode value assignable to the current block. In this case, the intra predictor may map the intra-prediction mode values of neighboring blocks (ie, block A and/or block B) to mode values assignable to the current block, and may then match the mapped The intra prediction mode corresponding to the mode value is determined as the MPM candidate corresponding to the adjacent blocks (ie, block A and/or block B). A method of mapping intra prediction mode values of adjacent blocks to mode values assignable to the current block is shown in Table 2 below as an embodiment.

[Table 2]

In the embodiment in Table 2, "value" indicates an intra prediction mode value of a neighboring block. Also, if the size of the neighboring block is the same or equal to the size of the current block and if the number of intra prediction modes assignable to the current block is 4, mapIntraPredMode3[value] may indicate that the intra prediction modes of the neighboring block are mapped to the mode value. If the size of the neighboring block is the same or equal to the size of the current block and if the number of intra prediction modes assignable to the current block is 18 or 35, mapIntraPredMode9[value] may indicate that the intra prediction modes of the neighboring block are mapped to the mode value.

For example, if the size of the current block is 64x64, the number of intra prediction modes assignable to the current block may be 4 (eg, intra prediction modes having a mode value ranging from 0 to 3). In this case, if the intra prediction mode value of the adjacent block exceeds 3, the intra predictor may map the intra prediction mode value of the adjacent block to a mode value less than or equal to 3 (ie, mapIntraPredMode3[value] ), as in the embodiment of Table 2, and the mapped mode values may thereafter be used as MPM candidates corresponding to neighboring blocks. For another example, if the size of the current block is 4x4, the number of intra prediction modes assignable to the current block may be 18 (ie, intra prediction modes with mode values ranging from 0 to 17). In this case, if the intra prediction mode value of the adjacent block exceeds 17, the intra predictor may map the intra prediction mode value of the adjacent block to a mode value less than or equal to 9 (ie, mapIntraPredMode9[value] ), as in the embodiment of Table 2, and the mapped mode values may thereafter be used as MPM candidates corresponding to neighboring blocks. For another example, if the number of intra prediction modes assignable to the current block is 18 and if the intra prediction mode value of the adjacent block exceeds 17, the intra predictor may map the intra prediction mode value of the adjacent block to Mode value less than or equal to 17.

Unlike the above-described embodiment, if the number of intra prediction modes assignable to the current block is a certain fixed value (ie, 35), the intra prediction modes of adjacent blocks (ie, block A and/or block B) among It is unlikely that the mode value of is greater than the maximum mode value that can be assigned to the current block. Therefore, in this case, the intra predictor cannot apply the mapping process described by the embodiment of Table 2 in the process of deriving MPM candidates.

In the embodiment of 810 in FIG. 8 , the MPM candidate A corresponding to the block A 813 and the MPM candidate B corresponding to the block B 816 can be derived through the above process. However, as described above, the number of MPM candidates constituting the MPM candidate list may be fixed to three. Therefore, the intra predictor may additionally determine two MPM candidates if MPM candidate A and MPM candidate B are the same, and may additionally determine one MPM candidate if MPM candidate A and MPM candidate B are not the same.

In one embodiment, if MPM candidate A is the same as MPM candidate B, the MPM candidates included in the MPM candidate list may be determined as follows. For example, if the MPM candidate A is the planar mode or the DC mode, the intra predictor may determine the planar mode, the DC mode, and the vertical mode as MPM candidates included in the MPM candidate list. Also, if MPM candidate A is neither the planar mode nor the DC mode, the intra predictor may determine MPM candidate A and two intra prediction modes having the most similar prediction directions to MPM candidate A to be included in the MPM candidate list MPM candidates in .

Also, if the MPM candidate A is not the same as the MPM candidate B, additional MPM candidates included in the MPM candidate list may be determined as follows. If neither MPM candidate A nor MPM candidate B is a plane mode, the plane mode may be determined as an additional MPM candidate. Also, if one of the MPM candidate A and the MPM candidate B is the planar mode and neither the MPM candidate A nor the MPM candidate B is the DC mode, the DC mode may be determined as an additional MPM candidate. Also, if one of the MPM candidate A and the MPM candidate B is the planar mode and the other is the DC mode, the vertical mode may be determined as an additional MPM candidate.

Meanwhile, the positions of neighboring blocks for deriving MPM candidates may be determined differently from the embodiment of 810 of FIG. 8 . 820 of FIG. 8 and 830 of FIG. 8 illustrate other embodiments of neighboring blocks for deriving MPM candidates.

For example, referring to 820 of FIG. 8 , a block A 823 located at the bottommost part among the left neighboring blocks located on the left side of the current block and a block located at the most right part among the upper neighboring blocks adjacent to the upper side of the current block Based on B826, the intra predictor can derive MPM candidates corresponding to the current block. For another example, referring to 830 of FIG. 8 , block A 833 in any part of the left adjacent block located to the left of the current block and block B 836 in any part of the upper adjacent block adjacent to the upper side of the current block On the basis of , the intra predictor can derive MPM candidates corresponding to the current block. Since the process of deriving MPM candidates for each case is similar to the embodiment of 810 of FIG. 8, it is omitted herein.

Hereinafter, for the convenience of explanation, it is assumed that the present invention, in the embodiment of 810 in FIG. 8 , is the uppermost block located in the left adjacent blocks located on the left side of the current block and the upper adjacent block located adjacent to the upper part of the current block. On the basis of the leftmost block in the block, MPM candidates corresponding to the current block can be derived. Also, for the convenience of explanation, it is assumed that the uppermost block located in the left adjacent blocks adjacent to the left side of the current block is referred to as a left candidate block (and/or block A) in the present invention, and is located adjacent to the current block The leftmost block among the upper adjacent upper adjacent blocks of is referred to as the upper candidate block (and/or block B). Also, the MPM candidate derived from block A is referred to as MPM candidate A, and the MPM candidate derived from block B is referred to as MPM candidate B. However, the following-derived embodiments are not limited thereto, and can also be equally or similarly applied to the case where the positions of adjacent blocks used to derive MPM candidates are different from the embodiments of 810 of FIG. 8 .

FIG. 9 is a diagram for explaining a line buffer in which an intra prediction mode is stored.

The plurality of LCUs of FIG. 9 are included in a picture, a slice, and/or a tile. Each square block of FIG. 9 corresponds to one LCU. In LCUxy of FIG. 9 (where x and y are integers greater than or equal to 0), then x may represent the row where the LCU is located, and y may represent the column where the LCU is located.

The intra predictor may perform prediction processing for each LCU shown in FIG. 9 , and may perform processing for the LCUs of FIG. 9 according to the raster scan order. For example, in the embodiment of FIG. 9, after processing for LCU line 910 corresponding to LCU0y is performed, in a left-to-right direction, processing for LCU line 920 corresponding to LCU1y may be performed.

Meanwhile, as described above in the embodiments of FIGS. 6 to 8 , on the basis of the intra prediction modes of neighboring blocks (and/or neighboring PUs) adjacent to the block, it can be derived from a block (and / or PU) corresponding MPM candidates. Because a PU is a unit belonging to an LCU, an intra prediction mode derived from one LCU must be stored in memory to process the LCU line located in the row immediately following the LCU line.

In this case, the encoder and decoder may store one intra-mode for each intra-mode storage unit. Here, the "intra mode storage unit" may mean the smallest unit by which the intra prediction mode is stored for use in prediction when performing intra prediction. For example, the intra mode storage unit may correspond to a block having a size of 4x4.

Likewise, if the intra prediction mode is stored in a block unit having a certain size, the intra prediction mode stored in the intra mode storage unit located in the lowest part in one LCU line can be stored in the buffer to The LCU line located in the row immediately following the LCU line is processed. In this case, in the present invention, the intra-mode memory cells stored in the buffer may constitute one line, and this is hereinafter referred to as an "intra-mode memory cell line".

It is assumed in FIG. 9 that the intra-mode storage unit corresponds to a block of size 4×4. Referring to FIG. 9 , for processing on the LCU line 920 corresponding to LCU1y, the intra prediction mode corresponding to LCU0y stored in the 4x4 block located in the lowest part in the LCU line 910 may be stored in a buffer. 930 of FIG. 9 indicates an intra-mode memory cell line composed of 4x4 blocks located at the lowermost portion of the LCD line 910 corresponding to LCU0y.

As in the above-described embodiments, the intra-prediction mode of the intra-mode storage unit line located at the bottom in the LCU line may be stored in a buffer to perform processing for the next LCU line. Likewise, a buffer that stores intra prediction modes belonging to an intra mode storage unit line may be referred to as an "intra mode line buffer". Hereinafter, the intra-mode line buffer may also be simply referred to as "line buffer" in the present invention.

FIG. 10 is a schematic diagram illustrating an embodiment of a method of deriving MPM candidates for blocks adjacent to a boundary of an LCU. In the embodiment of FIG. 10, the current block 1010 and neighboring blocks 1030 and 1040 adjacent to the current block may be corresponding blocks corresponding to one PU.

Referring to FIG. 10 , the current block 1010 may be a block located within one LCU line, and may be a block adjacent to the boundary 1020 of the LCU. Also, MPM candidates corresponding to the current block 1010 may be derived on the basis of the left candidate block A 1030 and the upper candidate block B 1040. In this case, the upper candidate block B may be a block belonging to an upper LCU line adjacent to the upper part of the current LCU line to which the current block belongs. Because LCUs are processed according to raster scan order, the intra-prediction modes of intra-mode storage unit lines (eg, lines composed of 4x4 sized blocks) located in the lower part of the LCU must be stored in the intra-mode line buffer for processing Blocks belonging to the current LCU line (eg, current block 1010).

Therefore, the size of the line buffer may increase in proportion to the width of the current picture to which the current block 1010 belongs. Since encoding/decoding performance decreases when the size of the line buffer increases, a method of deriving MPM candidates and a method of storing intra prediction modes are provided to reduce the size of the line buffer memory.

FIG. 11 is a schematic diagram illustrating an embodiment of a method for deriving MPM candidates according to the present invention.

In the embodiment of FIG. 11, the current block 1110 and neighboring blocks 1130 and 1140 adjacent to the current block may be corresponding blocks corresponding to one PU. Also, the numbers of the left candidate block A 1130 and the upper candidate block B 1140 may indicate an intra prediction mode value of each block. That is, in the embodiment of FIG. 11 , the intra prediction mode value of the left candidate block 1130 may be 10, and the intra prediction mode value of the upper candidate block 1140 may be 5. In this case, the intra prediction mode value 10 may correspond to the horizontal mode, and the intra prediction mode value 5 may correspond to one of a plurality of corner modes assignable to the current block 1110 .

Referring to FIG. 11 , the current block 1110 may be a block located within the current LCU, and may be a block adjacent to the upper boundary 1120 of the current LCU. In this case, the MPM candidate A derived from the left candidate block A 1130 may be the intra prediction mode of the left candidate block 1130 , that is, the horizontal mode corresponding to the mode value 10 .

However, since the upper candidate block B 1140 is a block belonging to an upper LCU adjacent to the upper part of the current LCU, it must be stored in the line buffer so that the intra prediction mode of the upper candidate block B 1140 is used as an MPM candidate. Therefore, by not using the intra prediction mode of the upper candidate block B 1140 as the MPM candidate of the current block 1110, the intra predictor may remove the line buffer for storing the intra prediction mode. This is because, if the intra prediction mode of the upper candidate block B1140 is not used as the MPM candidate of the current block 1110, the intra prediction mode of the upper candidate block B1140 is not necessarily stored in the line buffer.

That is, if the neighboring block used to derive the MPM candidate for the current block exists outside the LCU to which the current block belongs (and/or outside the boundary of the LCU to which the current block belongs), the intra predictor does not The intra prediction mode of the block is used as the MPM candidate corresponding to the adjacent block. As in the above-described embodiment, this may only be applied to the upper candidate block 1130, but the present invention is not limited thereto. therefore. This can also be applied only to the left candidate block or to both the upper and left candidate blocks. For example, if the current block is adjacent to the left boundary of the current LCU, the intra prediction mode of the left candidate block may not be used as the MPM candidate corresponding to the left candidate block.

If the neighboring blocks (ie, the left candidate block and/or the upper candidate block) for deriving the MPM candidate of the current block are located outside the current LCU, a certain intra prediction mode determined by a specific condition may be determined to be the same as the MPM candidates corresponding to neighboring blocks. For example, if the left candidate block is located outside the current LCU, the intra predictor may determine the MPM candidate A by assuming the intra prediction mode determined by a specific condition as the intra prediction mode of the left candidate block. Also, if the upper candidate block is located outside the current LCU, the intra predictor may determine the MPM candidate B by assuming the intra prediction mode determined by a specific condition as the intra prediction mode of the upper candidate block. It can be considered that a certain intra prediction mode determined by a specific condition is assigned to the intra prediction modes of adjacent blocks.

In one embodiment, the intra predictor may determine the DC mode as a relative block if the neighboring blocks used to derive the MPM candidate for the current block (ie, the left candidate block and/or the upper candidate block) are located outside the current LCU Intra prediction mode for neighboring blocks. In this case, the DC mode can be determined as the MPM candidate corresponding to the adjacent block. That is, the intra predictor can determine the DC mode as the MPM candidate corresponding to the adjacent block by assuming the intra prediction mode of the adjacent block as the DC mode.

For example, if the left candidate block is located outside the current LCU, the intra predictor may determine the DC mode as the intra prediction mode of the left candidate block. In this case, the DC mode may be determined as the MPM candidate A corresponding to the left candidate block. Also, if the upper candidate block is located outside the current LCU, the intra predictor may determine the DC mode as the intra prediction mode of the upper candidate block. In this case, the DC mode may be determined as the MPM candidate mode B corresponding to the upper candidate block.

If the upper candidate block is located outside the current LCU, the process of determining the DC mode as the intra prediction mode of the upper candidate block in one embodiment is described as follows.

- The candidate intra prediction mode CandIntraPredModeN is derived as follows.

- if N equals B and yB1 is less than

((yB>>Log2CtbSizeY)<<Log2CtbSizeY),

Then intraPredModeB is set equal to Intra_DC.

Here, candIntraPredModeN may indicate MPM candidates. In this case, if N is A, candIntraPredModeN may correspond to the MPM candidate A corresponding to the left candidate block (ie, block A). Also, if N is B, candIntraPredModeN may correspond to the MPM candidate B corresponding to the upper candidate block (ie, block B). Also, yB may represent the y-coordinate of the pixel located in the uppermost left portion of the current block, and log2CtbSizeY may represent the logarithmic value (ie, the size in the y-axis direction) of the height of the LCU to which the current block belongs. Also, intraPredModeB may represent the intra prediction mode of the upper candidate block, and Intra_DC may represent the DC mode.

For another example, the intra predictor may determine the plane mode as adjacent if the adjacent blocks used to derive the MPM candidate for the current block (ie, the left candidate block and/or the upper candidate block) are located outside the current LCU. Intra prediction mode for the block. In this case, the plane mode can be determined as the MPM candidate corresponding to the adjacent block. That is, the intra predictor can determine the plane mode as the MPM candidate corresponding to the adjacent block by assuming the intra prediction mode of the adjacent block as the plane mode.

For example, if the left candidate block is located outside the current LCU, the intra predictor may determine the plane mode as the intra prediction mode of the left candidate block. In this case, the plane mode may be determined as the MPM candidate A corresponding to the left candidate block. Also, if the upper candidate block is located outside the current LCU, the intra predictor may determine the plane mode as the intra prediction mode of the upper candidate block. In this case, the plane mode may be determined as the MPM candidate B corresponding to the upper candidate block.

Referring to FIG. 11 , the upper candidate block B 1140 may be a block belonging to an upper LCU adjacent to the upper portion of the current LCU. Therefore, the intra predictor cannot use the intra prediction mode of the upper candidate block B1140, which is the corner mode corresponding to the mode value 5, as an MPM candidate. In this case, for example, the intra predictor may determine the plane mode as the intra prediction mode of the upper candidate block B1140. In this case, the MPM candidate B corresponding to the upper candidate block B1140 may be determined as the plane mode. For example, the mode value of the planar mode may be 0 here.

Meanwhile, neighboring blocks (eg, left candidate block and/or upper candidate block) for deriving MPM candidates of the current block may not be available. For example, if a neighboring block (ie, the left candidate block and/or the upper candidate block) is located outside the current picture to which the current block belongs (and/or outside the current picture boundary to which the current block belongs), the neighboring block may Corresponds to unavailable blocks. Also, if a neighboring block (ie, the left candidate block and/or the upper candidate block) is located outside the current slice to which the current block belongs (and/or outside the current slice boundary to which the current block belongs), the neighboring block may Corresponds to unavailable blocks.

Furthermore, even in the case where the neighboring blocks themselves (eg, the left candidate block and/or the upper candidate block) used to derive the MPM candidate of the current block are unavailable, the prediction mode of the neighboring block may not be the intra mode. In this case, adjacent blocks may not include valid intra prediction mode information.

As described above, if adjacent blocks (ie, the left candidate block and/or the upper candidate block) are not available or if the prediction mode of the adjacent block is not intra mode, intra prediction may convert a certain frame determined by certain conditions Intra prediction modes are determined as MPM candidates corresponding to adjacent blocks.

In one embodiment, the intra predictor may determine the DC mode if adjacent blocks (ie, the left candidate block and/or the upper candidate block) are unavailable or if the prediction mode of the adjacent block is not intra mode is the MPM candidate corresponding to the adjacent block. That is, the intra predictor can determine the DC mode as the MPM candidate corresponding to the adjacent block by assuming the intra prediction mode of the adjacent block as the DC mode. For example, if the left candidate block is unavailable or not an intra block, the intra predictor may determine the DC mode as the MPM candidate A corresponding to the left candidate block. Also, if the upper candidate block is unavailable or not an intra block, the intra predictor may determine the DC mode as the MPM candidate B corresponding to the upper candidate block.

If the neighboring block (ie, the left candidate block and/or the upper candidate block) is not available or if the prediction mode of the neighboring block is not intra mode, then in one embodiment the DC mode is determined to correspond to the neighboring block The processing of MPM candidates is expressed as follows.

- The candidate intra prediction mode CandIntraPredModeN is derived as follows.

- CandIntraPredModeN if availableN is equal to FALSE

is set equal to Intra_DC.

Otherwise, if PredMode[xBN][yBN] is not equal to MODE_INTRA

Then CandIntraPredModeN is set equal to Intra_DC.

Here, availableN may indicate whether adjacent blocks (ie, the left candidate block and/or the upper candidate block) are available. In this case, if N is A, availableN may indicate whether the left candidate block (ie, block A) is available, and if N is B, availableN may indicate whether the upper candidate block (ie, block B) is available of. Also, PredMode[xBN][yBN] may indicate the prediction mode of neighboring blocks (i.e., the left candidate block and/or the upper candidate block), and MODE_INTRA may indicate the intra mode.

As another example, the intra predictor may determine the plane mode to be the same as if the adjacent block (ie, the left candidate block and/or the upper candidate block) is unavailable or if the prediction mode of the adjacent block is not the intra mode. The MPM candidate corresponding to the adjacent mode. That is, the intra predictor can determine the plane mode as the MPM candidate corresponding to the adjacent block by assuming the intra prediction mode of the adjacent block as the plane mode. For example, if the left candidate block is unavailable or not an intra block, the intra predictor may determine the plane mode as the MPM candidate A corresponding to the left candidate block. Also, if the upper candidate block is unavailable or not an intra block, the intra predictor may determine the plane mode as the MPM candidate B corresponding to the upper candidate block.

In another embodiment, if the neighboring block (ie, the left candidate block and/or the upper candidate block) is unavailable or if the prediction mode of the neighboring block is not intra mode, the intra predictor may The intra prediction modes of different neighboring blocks adjacent to the block are determined as MPM candidates corresponding to the neighboring blocks. For example, if the left candidate block is unavailable or not an intra block, the intra predictor may determine the intra prediction mode of a different neighboring block adjacent to the left candidate block as the MPM candidate A corresponding to the left candidate block. Also, if the upper candidate block is unavailable or not an intra block, the intra predictor may determine intra prediction modes of different neighboring blocks adjacent to the upper candidate block as MPM candidate B corresponding to the upper candidate block.

In the above-described embodiment of FIG. 11 and its subsequent figures, at the position of the left candidate block (ie, whether it exists outside the picture, slice and/or the LCU to which the current block belongs) and/or the position of the left candidate block The MPM candidate A can be determined on the basis of the prediction mode, and on the basis of the position of the upper candidate block (ie, whether it exists outside the picture, slice and/or the LCU to which the current block belongs) and/or the prediction mode, can be Determine MPM candidate B. The above-described embodiment of FIG. 11 and its subsequent figures may be applied independently or may be combined in a selected manner to be applied to the method of deriving MPM candidates.

For example, if the upper candidate block is located outside the current LCU, the intra predictor may determine the intra prediction mode of the upper candidate block as DC mode (or plane mode). In this case, the MPM candidate B corresponding to the upper candidate block may be determined as the DC mode (or the plane mode). Also, if the upper candidate block is unavailable or not an intra block, the intra predictor may determine the DC mode (or plane mode) as the MPM candidate B corresponding to the upper candidate block.

Also, if the left candidate block is unavailable or not an intra block, the intra predictor may determine the DC mode (or plane mode) as the MPM candidate B corresponding to the left candidate block. However, if the left candidate block is located outside the current LCU, the intra prediction mode of the left candidate block may be directly determined as the MPM candidate A, unlike the upper candidate block.

In this case, since the intra prediction mode of the previous candidate block is not used as the MPM candidate of the current block, the intra predictor can derive the MPM candidate without using the intra mode line buffer. Therefore, the line buffer can be removed in the above-described embodiment.

FIG. 12 is a schematic diagram illustrating an embodiment of a method of deriving MPM candidates on the basis of intra mode storage units. In the embodiment of FIG. 12, the current block 1210, the left candidate block A 1220, and the upper candidate block B 1230 may be corresponding blocks corresponding to one PU.

As described above, the MPM candidate of the current block 1210 may be derived on the basis of the intra prediction mode of the left candidate block A 1220 and the intra prediction mode of the upper candidate block B 1230 . In this case, the intra prediction mode of the left candidate block A 1220 and the intra prediction mode of the upper candidate block B 1230 must be stored in the buffer to process the current block 1210 . Specifically, if the current block 1210 is adjacent to the upper boundary of the current LCU including the current block 1210, the intra prediction mode of the upper candidate block B 1230 may be stored in the intra mode line moderator. In this case, one intra prediction mode may be stored in the buffer for each intra mode storage unit. Assume that in the embodiment of FIG. 12, the intra-mode storage unit corresponds to a block of size 4x4.

The intra prediction mode of the left candidate block A 1220 can be obtained from the uppermost block in the 4×4 size block (ie, the intra mode storage unit) adjacent to the left side of the current block 1210 . In this case, the 4x4-sized block through which the intra prediction mode is obtained may be the intra mode storage unit belonging to the left candidate block A1220. In the embodiment of FIG. 12, since the left candidate block A 1220 has a size of 4x4, the intra-mode storage unit may have the same size as the left candidate block A 1220.

Also, from the leftmost block 1240 located in the 4x4-sized block (ie, the intra mode storage unit) adjacent to the upper part of the current block 1210, the intra prediction mode of the upper candidate block B 1230 can be obtained. In this case, the 4x4-sized block 1240 may be an intra-mode storage unit belonging to the upper candidate block B 1230 . In the embodiment of FIG. 12 , since the upper candidate block B1230 has a size of 8×8, the intra-mode storage unit may have a smaller size than the upper candidate block B1230. In this case, the 4×4-sized block 1240 through which the intra prediction mode is obtained may be a block located at the lowest part of the upper candidate block B 1230 .

As described in the above embodiment, if the intra mode storage unit is a 4x4 sized block, one intra prediction mode may be stored for each 4x4 sized block. In this case, since the load of the buffer (and/or line buffer) in which the intra prediction mode is stored may be large, the encoder and decoder may store one intra prediction mode with respect to multiple 4x4 sized blocks, Thereby the buffer (and/or line buffer) size can be reduced.

For example, the encoder and decoder may store one intra prediction mode (and/or line buffer) for each of two 4x4 sized blocks. In this case, the size of the intra-mode storage unit may correspond to a block of size 8x4. Furthermore, the number of intra prediction modes stored in the buffer (and/or line buffer) and the size of the buffer (and/or line buffer) can be reduced by 1/2. Therefore, this method of storing intra prediction modes may also be referred to as "2:1 buffer (and/or line buffer) compression" or "2:1 intra mode compression".

As another example, the encoder and decoder may store one intra prediction mode in the buffer (and/or line buffer) for each of four 4x4 sized blocks. In this case, the size of the intra-mode storage unit may correspond to a block of size 16x4. Also, the number of intra prediction modes stored in the buffer (and/or line buffer) and the size of the buffer (and/or line buffer) can be reduced by 1/4. Therefore, this method of storing intra prediction modes may also be referred to as "4:1 buffer (and/or line buffer) compression" or "4:1 intra mode compression".

As mentioned above, if the size of the intra-mode storage unit is extended, the encoder and decoder store in the buffers (and/or line buffers) in the 4x4-sized block belonging to the intra-mode storage unit only the memory allocated to Intra prediction mode for a block. In this case, the encoder and the decoder may use only one intra prediction mode to derive MPM candidates in a plurality of intra prediction modes included in one intra mode storage unit. That is, when deriving MPM candidates on the basis of the intra prediction modes of blocks located outside the current LCU (ie, the LCU to which the current block belongs), the encoder and decoder only use buffers stored in buffers with reduced sizes (and and/or the compressed intra prediction mode in the line buffer).

Although the line buffer is removed in the above-described embodiment of FIG. 11, if an intra-mode compression scheme (and/or a line buffer compression scheme) is used, the line buffer may not be removed and only the line buffer's size is reduced. Therefore, since the intra prediction modes of adjacent blocks are considered when deriving MPM candidates, the intra prediction modes can be predicted more accurately than the embodiment of FIG. 11 .

13 is a schematic diagram illustrating an embodiment of a 2:1 line buffer compression scheme.

1310 to 1340 of FIG. 13 indicate corresponding intra-mode memory cell lines. As described above with reference to FIG. 9 , the intra prediction mode stored in the intra mode storage unit may be used to process lower LCU lines adjacent to the lower portion of the LCU line to which the intra mode storage unit line belongs. If the current block is a block belonging to the lower LCU line, the upper candidate block corresponding to the current block may be a block belonging to an LCU line adjacent to the upper part of the lower LCU line. In this case, the intra predictor may use the intra prediction mode stored in the intra mode storage unit line to derive MPM candidates for the current block.

Each square block of FIG. 13 indicates a 4x4 size block. Furthermore, in the embodiment of FIG. 13 , line 1350 indicates a line corresponding to the boundary of an 8x8 size block (hereinafter, referred to as "8x8 block boundary"), and line 1360 indicates a line corresponding to the boundary of a 16x16 size block line (hereinafter, referred to as "16x16 block boundary"). Although the boundaries of 16x16 sized blocks may correspond to the boundaries of 8x8 sized blocks, line 1360 may also be referred to as a 16x16 block boundary in the embodiment of FIG. 13 .

Referring to FIG. 13, one intra-mode memory cell line may be composed of a plurality of intra-mode memory cells. In this case, the intra-mode storage unit may consist of two blocks of size 4x4, and may have a size of 8x4. That is, the encoder and decoder can store one intra prediction mode for each of the two 4x4 blocks. In this case, the memory size for storing the intra prediction mode can be reduced by 1/2. In the embodiment of FIG. 13, each intra-mode memory cell constituting a line of intra-mode memory cells may be located between an 8x8 block boundary and an adjacent 16x16 block boundary.

In the 2:1 line buffer compression scheme, only intra prediction modes corresponding to one block may be stored in a line buffer between two 4x4 sized blocks constituting one intra mode storage unit line. In this case, the upper candidate block referenced by the current block to derive the MPM candidate may include a 4x4 block of the intra prediction mode that is not stored in the line buffer. In this case, the encoder and/or decoder may use the intra prediction mode of another 4x4 block belonging to the same intra mode storage unit as the 4x4 block (ie, the block in which the intra prediction mode is stored) to Derive MPM candidates.

Referring to 1310 of FIG. 13 , the encoder and the decoder may store only the intra prediction mode of the block located on the left into the line buffer between two 4×4 sized blocks belonging to the intra mode storage unit. In this case, the block referenced by the current block to derive the MPM candidate may be a 4x4-sized block located on the right side in the intra-mode storage unit (eg, a block included in the upper candidate block corresponding to the current block) . In this case, the encoder and the decoder may use the intra prediction mode of the block located in the left side in the intra mode storage unit to derive the MPM candidate.

That is, in the embodiment of 1310 of FIG. 13, the 4x4 block on the left in the intra-mode storage unit may represent two 4x4 blocks belonging to the intra-mode storage unit. Also, the 4x4 blocks located on the right side in the intra mode storage unit may share the intra prediction mode of the 4x4 blocks located on the left side.

Referring to 1320 of FIG. 13 , the encoder and the decoder store only the intra prediction mode of the block located on the right into the line buffer between two 4×4 sized blocks belonging to the intra mode storage unit. In this case, the block referenced by the current block to derive the MPM candidate may be a 4x4 size block located on the left side in the intra mode storage unit (eg, a block included in the upper candidate block corresponding to the current block) . In this case, the encoder and the decoder may use the intra prediction mode of the block located on the right in the intra mode storage unit to derive the MPM candidate.

That is, in the embodiment of 1320 of FIG. 13, the 4x4 block located to the right in the intra-mode storage unit may represent two 4x4 blocks belonging to the intra-mode storage unit. Also, the 4x4 blocks located on the left side of the intra mode storage unit may share the intra prediction mode of the 4x4 blocks located on the right side.

Referring to 1330 of FIG. 13 , the encoder and the decoder may store only intra prediction modes of blocks adjacent to a 16x16 block boundary 1360 between two 4x4 sized blocks belonging to the intra mode storage unit. In this case, the block referenced by the current block to derive the MPM candidate may be a block of size 4x4 adjacent to the 8x8 block boundary 1350 in the intra mode storage unit (eg, including the upper candidate block corresponding to the current block) in the block). In this case, the encoder and decoder may use the intra prediction mode of the block adjacent to the 16x16 block boundary 1360 in the intra mode storage unit to derive MPM candidates.

That is, in the embodiment of 1330 of Figure 13, a 4x4 block adjacent to a 16x16 block boundary 1360 in the intra-mode storage unit may represent two 4x4 blocks belonging to the intra-mode storage unit. Also, the 4x4 blocks adjacent to the 8x8 block boundary 1350 in the intra mode storage unit may share the intra prediction mode of the 4x4 blocks adjacent to the 16x16 block boundary 1360.

Referring to 1340 of FIG. 13 , the encoder and the decoder may store only intra prediction modes of blocks adjacent to the 8x8 block boundary 1350 between two 4x4 sized blocks belonging to the intra mode storage unit. In this case, the block referenced by the current block to derive the MPM candidate may be a 4x4 sized block adjacent to the 16x16 block boundary 1360 in the intra mode storage unit (eg, including the upper candidate block corresponding to the current block) in the block). In this case, the encoder and decoder may use the intra prediction mode of the block adjacent to the 8x8 block boundary 1350 in the intra mode storage unit to derive MPM candidates.

That is, in the embodiment of 1340 of Figure 13, the 4x4 blocks adjacent to the 8x8 block boundary 1350 in the intra-mode storage unit may represent two 4x4 blocks belonging to the intra-mode storage unit. In addition, the 4x4 block adjacent to the 16x16 block boundary 1360 in the intra mode storage unit may share the intra prediction mode of the 4x4 block adjacent to the 8x8 block boundary 1350.

14 is a schematic diagram illustrating an embodiment of a 4:1 line buffer compression scheme.

1410 to 1440 of FIG. 14 indicate corresponding intra-mode memory cell lines. As described above with reference to FIG. 9 , the intra prediction mode stored in the intra mode storage unit may be used to process lower LCU lines adjacent to the lower portion of the LCU line to which the intra mode storage unit line belongs. If the current block is a block belonging to the lower LCU line, the upper candidate block corresponding to the current block may be a block belonging to an LCU line adjacent to the upper part of the lower LCU line. In this case, the intra predictor may use the intra prediction mode stored in the intra mode storage unit line to derive MPM candidates for the current block.

Each square block of FIG. 14 indicates a 4x4 size block. Furthermore, in the embodiment of FIG. 14, line 1470 indicates a line corresponding to the boundary of a 16x16-sized block (hereinafter, referred to as "16x16 block boundary"), and line 1480 indicates a line corresponding to a boundary of a 32x32-sized block line (hereinafter, referred to as "32x32 block boundary"). Although the boundaries of 32x32 sized blocks may correspond to the boundaries of 16x16 sized blocks, line 1480 may also be referred to as a 32x32 block boundary in the embodiment of FIG. 14 .

Referring to FIG. 14, one intra-mode memory cell line may be composed of a plurality of intra-mode memory cells. In this case, the intra-mode storage unit may consist of four blocks of size 4x4, and may have a size of 16x4. That is, the encoder and decoder can store one intra prediction for each of the four 4x4 blocks. In this case, the size of the memory for storing the intra prediction mode can be reduced by 1/4. In the embodiment of FIG. 14, each intra-mode memory cell constituting a line of intra-mode memory cells may be located between two 16x16 block boundaries adjacent to each other.

Hereinafter, among the four 4x4-sized blocks constituting the intra-mode storage unit in the embodiment of FIG. 14 , the leftmost block is referred to as the first 4x4 block, and the 4x4 adjacent to the first 4x4 block The block of size is called the second 4x4 block. Furthermore, in the embodiment of FIG. 14 , the block of size 4x4 adjacent to the right side of the second 4x4 block in the intra-mode storage unit is referred to as the third 4x4 block, and is opposite to the right side of the third 4x4 block. The adjacent 4x4 size block is called the fourth 4x4 block.

In the 4:1 line buffer compression scheme, only an intra prediction mode corresponding to one block can be stored in a line buffer in four 4x4 sized blocks constituting one intra mode storage unit line. In this case, the upper candidate block referenced by the current block to derive the MPM candidate may include a 4x4 block whose intra prediction mode is not stored in the line buffer. In this case, the encoder and/or decoder may use intra prediction modes stored in line buffers in intra prediction modes of other 4x4 blocks belonging to the same intra mode storage unit as the 4x4 block to derive MPM candidate.

Referring to 1410 of FIG. 14 , the encoder and the decoder may store only the intra prediction mode of the first 4x4 block among the four 4x4 sized blocks belonging to the intra mode storage unit. In this case, the block referenced by the current block to derive the MPM candidate may be another 4x4 block (eg, the second 4x4 block, the third 4x4 block, or the fourth 4x4 block) in addition to the first 4x4 block. In this case, the encoder and decoder may use the intra prediction mode of the first 4x4 block in the intra mode storage unit to derive MPM candidates.

That is, in the embodiment of 1410 of Figure 14, the first 4x4 block in the intra-mode storage unit may represent four 4x4 blocks belonging to the intra-mode storage unit. Also, the second 4x4 block, the third 4x4 block and the fourth 4x4 block in the intra mode storage unit may share the intra prediction mode of the first 4x4 block.

Referring to 1420 of FIG. 14 , the encoder and the decoder may store only the intra prediction mode of the second 4x4 block to the line buffer in four 4x4 sized blocks belonging to the intra mode storage unit. In this case, the block referenced by the current block to derive the MPM candidate may be another 4x4 block (eg, the first 4x4 block, the third 4x4 block, or the fourth 4x4 block) in addition to the second 4x4 block. In this case, the encoder and decoder may use the intra prediction mode of the second 4x4 block in the intra mode storage unit to derive MPM candidates.

That is, in the embodiment of 1420 of Figure 14, the second 4x4 block in the intra-mode storage unit may represent four 4x4 blocks belonging to the intra-mode storage unit. Also, the first 4x4 block, the third 4x4 block, and the fourth 4x4 block in the intra mode storage unit may share the intra prediction mode of the second 4x4 block.

Referring to 1430 of FIG. 14 , the encoder and the decoder may store only the intra prediction mode of the third 4x4 block among the four 4x4 sized blocks belonging to the intra mode storage unit. In this case, the block referenced by the current block to derive the MPM candidate may be another 4x4 block (eg, the first 4x4 block, the second 4x4 block, or the fourth 4x4 block) in addition to the third 4x4 block. In this case, the encoder and decoder may use the intra prediction mode of the third 4x4 block in the intra mode storage unit to derive MPM candidates.

That is, in the embodiment of 1430 of Figure 14, the third 4x4 block in the intra-mode storage unit may represent four 4x4 blocks belonging to the intra-mode storage unit. Also, the first 4x4 block, the second 4x4 block and the fourth 4x4 block in the intra mode storage unit may share the intra prediction mode of the third 4x4 block.

Referring to 1440 of FIG. 14 , the encoder and the decoder may store intra prediction modes of only four 4x4 blocks out of four 4x4 sized blocks belonging to the intra mode storage unit. In this case, the block referenced by the current block to derive the MPM candidate may be another 4x4 block (eg, the first 4x4 block, the second 4x4 block, or the third 4x4 block) in addition to the fourth 4x4 block. In this case, the encoder and decoder may use the intra prediction mode of the fourth 4x4 block in the intra mode storage unit to derive MPM candidates.

That is, in the embodiment of 1440 of FIG. 14, the fourth 4x4 block in the intra-mode storage unit may represent four 4x4 blocks belonging to the intra-mode storage unit. Also, the first 4x4 block, the second 4x4 block and the third 4x4 block in the intra mode storage unit may share the intra prediction mode of the fourth 4x4 block.

Referring to 1450 of FIG. 14 , the encoder and the decoder may store only intra prediction modes of blocks adjacent to a 32×32 block boundary 1480 among four 4×4 sized blocks belonging to the intra mode storage unit. In this case, the block referenced by the current block to derive the MPM candidate may be a 4x4 block that is not adjacent to the 32x32 block boundary 1480 in the intra mode storage unit. In this case, the encoder and decoder may use the intra prediction mode of the block adjacent to the 32x32 block boundary 1480 in the intra mode storage unit to derive MPM candidates.

That is, in the embodiment of 1450 of Figure 14, the 4x4 blocks adjacent to the 32x32 block boundary 1480 in the intra-mode storage unit may represent four 4x4 blocks belonging to the intra-mode storage unit. Also, three 4x4 blocks not adjacent to the 32x32 block boundary 1480 in the intra mode storage unit may share the intra prediction mode of the 4x4 block adjacent to the 32x32 block boundary 1480.

Referring to 1460 of FIG. 14 , the encoder and the decoder may store only intra prediction modes of blocks adjacent to a 16×16 block boundary 1470 among four 4×4 sized blocks belonging to the intra mode storage unit. In this case, the block referenced by the current block to derive the MPM candidate may be a 4x4 block that is not adjacent to the 16x16 block boundary 1470 in the intra mode storage unit. In this case, the encoder and decoder may use the intra prediction mode of the block adjacent to the 16x16 block boundary 1470 in the intra mode storage unit to derive MPM candidates.

That is, in the embodiment of 1460 of Figure 14, the 4x4 blocks adjacent to the 16x16 block boundary 1470 in the intra-mode storage unit may represent four 4x4 blocks belonging to the intra-mode storage unit. Also, three 4x4 blocks not adjacent to the 16x16 block boundary 1470 in the intra mode storage unit may share the intra prediction mode of the 4x4 block adjacent to the 16x16 block boundary 1470.

In 1410 to 1460 of FIG. 14 described above, an embodiment of a block in which the intra prediction mode is stored in four 4x4 blocks belonging to the intra mode storage unit is described. However, the present invention is not limited to this, and thus 4:1 line buffer compression can also be equally or similarly applied to the position of the block in which the intra prediction mode is stored is determined as described in 1410 to 1460 of FIG. 14 . those different situations.

Meanwhile, the above buffer (and/or line buffer) compression scheme described in the embodiments of FIGS. 12 to 14 may be applied to all blocks to be encoded/decoded in intra mode as one example, or as another The example applies only to blocks adjacent to the LCU boundary. As another example, a buffer (and/or line buffer) compression scheme may be applied to both the left and upper candidate blocks used to derive MPM candidates, or may be applied only to the space between the left and upper candidate blocks a block.

These embodiments may be applied independently or combined in a selective manner to be applied to the process of deriving MPM candidates.

For example, encoders and decoders cannot apply the aforementioned buffers (and/or line buffers) to left candidate blocks. In this case, the intra prediction mode corresponding to the left candidate block may be stored in a buffer (and/or a line buffer) in a 4x4 block unit. Furthermore, encoders and decoders may apply the above-described buffers (and/or line buffers) to the upper candidate block. In this case, with regard to the upper candidate block, one intra prediction mode may be stored for each of the two (or four) 4x4 blocks. In this case, the size of the buffer (and/or line buffer) for storing the intra prediction mode can be reduced by 1/2 (or 1/4).

Although the above-described exemplary system has been described on the basis of a flowchart in which steps or blocks are listed in order, the steps of the invention are not limited to a certain order. Thus, with respect to the above, a certain step may be performed in different steps or in a different order or simultaneously. Furthermore, those skilled in the art will understand that the steps of these flowcharts are not exclusive. Rather, another step may be included, or one or more steps may be deleted within the scope of the present invention.

The above-described embodiments include various exemplary aspects. While not all possible combinations for representing the various aspects have been described, it will be apparent to those skilled in the art that other combinations are possible. Therefore, all substitutions, modifications and changes should fall within the spirit and scope of the claims of the present invention.

Claims (6)

1. A decoding apparatus that performs intra prediction, comprising:

a predictor for setting a first candidate intra prediction mode related to a left candidate block adjacent to a left side of a current block based on availability of the left candidate block or an intra prediction mode, for setting a second candidate intra prediction mode related to an upper candidate block adjacent to an upper side of the current block based on availability of the upper candidate block or the intra prediction mode, for generating a candidate mode list including a plurality of candidate intra prediction modes based on the first candidate intra prediction mode and the second candidate intra prediction mode, for determining one candidate intra prediction mode as an intra prediction mode of the current block among the plurality of candidate intra prediction modes constituting the candidate mode list, and for performing intra prediction based on the determined intra prediction mode to generate a prediction block including prediction pixels,

wherein the left candidate block is adjacent to an uppermost portion of the left side of the current block,

wherein the upper candidate block is adjacent to a leftmost portion of an upper side of the current block,

wherein the first candidate intra prediction mode is set to a DC mode when the left candidate block is unavailable, and

wherein the second candidate intra-prediction mode is set to a DC mode when the upper candidate block is unavailable.

2. The apparatus of claim 1, wherein the first and second electrodes are disposed on opposite sides of the housing,

wherein the predictor determines that the left candidate block is unavailable and sets the first candidate intra prediction mode to a DC mode when the left candidate block is located outside a coding tree block to which the current block belongs,

wherein the predictor determines that the up-candidate block is unavailable and sets the second candidate intra prediction mode to a DC mode when the up-candidate block is located outside the coding tree block to which the current block belongs.

3. The apparatus of claim 1, wherein the first and second electrodes are disposed on opposite sides of the housing,

wherein the predictor determines that the left candidate block is unavailable and sets the first candidate intra prediction mode to a DC mode when the left candidate block is located outside a current picture to which the current block belongs, and

wherein the predictor determines that the candidate block is unavailable and sets the second candidate intra-prediction mode to DC mode when the candidate block is outside the current picture.

4. The apparatus of claim 1, wherein the first and second electrodes are disposed on opposite sides of the housing,

wherein the predictor determines that the left candidate block is unavailable and sets the first candidate intra-prediction mode to DC mode when the left candidate block is outside a current slice, and

wherein the predictor determines that the candidate block is unavailable and sets the second candidate intra-prediction mode to DC mode when the candidate block is outside the current slice.

5. The apparatus of claim 1, wherein the first and second electrodes are disposed on opposite sides of the housing,

wherein when the prediction mode of the left candidate block is not an intra mode, the predictor determines that the left candidate block is unavailable and sets the first candidate intra prediction mode to a DC mode, and

wherein when the prediction mode of the candidate block is not the intra mode, the predictor determines that the candidate block is unavailable and sets the second candidate intra prediction mode to a DC mode.

6. The apparatus of claim 1, wherein the first and second electrodes are disposed on opposite sides of the housing,

wherein the predictor determines that the left candidate block is unavailable and sets the first candidate intra mode to DC mode when the left candidate block is located outside a current tile to which the current block belongs, and

wherein the predictor determines that the up-candidate block is unavailable and sets the second candidate intra-prediction mode to DC mode when the up-candidate block is located outside a current tile to which the current block belongs.

Images